In concert with Projects 1-3. this revised Project 4 probes the effects of fragmented eariy life experience on neuronal network structure and function using magnetic resonance brain imaging (MRI) of rats (with Project 1) and humans (with Projects 2-3). The results will be integrated with parameters generated by the other projects to accomplish the Center's goal of generating predictive models and markers of adolescent mental vulnerabilities. We will start by examining whether fragmented eariy-life experience influences the structure of brain regions and networks that are salient to cognitive and emotional functions. We then define a trajectory of these structural and functional alterations with development, and their correlation with cognitive and emotional behavior, resulting in potential biomarkers of vulnerability to overt cognitive and emotional pathology. The goal of this project is to employ MRI-derived measures to establish predictors of regional brain connectivity (as well as structural changes) that best correlate with developmental and cognitive vulnerabilities as a function of early life exposure to fragmented maternal signals. This novel analysis will identify a series of pathological changes that occur at varying intervals in the pre-symptomatic period that might guide the timing of future interventions and provide insights into intrinsic compensatory mechanisms. Its significance derives from the crucial importance of the clinical hypothesis: that fragmented patterns of sensory input modulate the function and connectivity of brain networks. The Project innovation stems from (a) The concept of developmentally evolving neuronal networks as the target of fragmented/unpredictable maternal input, (b) from the use of novel methodologies (e.g.. Structural Equation Modeling); (c) from the use of analysis of distributed hippocampal connectivity using multiple modalities across species, and (d) from inclusion of MRI parameters in multivariate models orchestrated by the Computational Core, to generate potentially predictive models for adolescent vulnerabilities.